基于神经网络的6-SPS并联机器人正运动学精确求解

位置正解是并联机器人机构应用的基础。探讨了人工神经网络在并联机器人机构位置正解求解中的应用。采用BP网络，利用位置逆解结果，通过训练学习，实现操作手从关节变量空间到工作变量空间的非线性映射；从而求得6-SPS并联机器人运动学正解。为提高正解结果精度，采用迭代计算进行误差补偿。给出了一种并联机器人操作手的仿真实例。计算结果表明，该法迭代次数少，计算精度高且计算速度接近机器人实时控制的要求。

An accurate solution for forward kinematics of 6-SPS Stewart platform based on neural network

The nonlinear mapping from the joint variable space to the operation variable space for the platform is accomplished with BP neural network after training and learning. The training data comes from the inversed kinematical model of the platform. To improve the accuracy of the solution, an iterative approach is used to compensate for the offset error. Numerical results have shown that the accurate solutions can be obtained by performing only a few iteration steps and the computation speed can meet the requirements of the robot's real time control system.